Taxes put bite on middle-class families

Eric Lobbecke

"MIDDLE-CLASS welfare" is not a descriptor; it is a rhetorical device. After all, who but the profligate could defend providing welfare to those who should be able to look after themselves?

Yet however effective that debating trick may be, the term is little more than an epithet in search of reality.

In fact, far from spraying largesse on the undeserving rich, the striking feature of our tax/transfer system is just how progressive it is. Strict means-testing of welfare payments ensures the poorest 20 per cent of Australians receive over 40 per cent of income transfers; the richest 20 per cent receive barely 3 per cent.

The poorest fifth therefore receive 12 times as much in cash benefits as the richest fifth, by far the highest ratio in the advanced economies. Our welfare system is consequently exceptionally redistributive: for each dollar spent on transfers, we reduce income inequality by 50 per cent more than Denmark or Norway and by two to three times more than Italy or France.

At the same time, the burden of income taxation - our largest revenue source - falls almost entirely on middle and upper-income earners. And the interaction of means-tested benefits with the income tax results in very high tax rates for many middle income earners, and especially for those in families.

Taking, for example, a family with a primary earner whose income is about $65,000, the effective marginal tax rate on a second earner's income can be as high as 65 to 70 per cent. And those high marginal rates translate into high average rates of tax: while a single earner in the bottom fifth of the income distribution receives a net transfer equal to 20 per cent of their income, a second earner with a personal income no higher would, on average, pay 25 to 30 per cent of (typically) her income in tax.

But the mere fact that there is so little middle-class welfare cuts no ice with its critics. Their complaints centre on family benefits; listening to their protests, one would think those benefits were little more than a subsidy to the well off.

Viewed as an empirical statement, that claim is absurd. Family benefits are steeply means tested: a couple with two children of primary school age in the lower fifth of the earned income distribution would qualify for $17,000 in family benefits and pay virtually no income tax; in contrast, a middle income family would receive $3000 in family benefits while paying $20,000 in tax.

Sure, the means testing could be made even more stringent; but that would lead to higher effective marginal tax rates, discouraging labour force participation and increasing the weight that would have to be placed on costly-to-administer (and often ineffective) "work availability" requirements.

But those facts too do nothing to silence the critics. Rather, they would go even further, disputing whether families with children should receive any benefits at all, with the possible exception of those that would otherwise be destitute.

That argument seems to rest on an almost complete misunderstanding of the principles underpinning tax design. Central among them is the goal of horizontal equity: that people should be taxed according to their capacity to pay, so two people should only pay the same amount of tax if their capacity to pay is the same.

Judged in terms of capacity to pay, a family with two children and an income of $80,000 is obviously different from a couple with the same income and no children: the first has an income per person half that of the second. Ignoring for the moment cost differences between children and adults, the first should therefore pay no more tax, in net terms, than would four people with an income of $20,000, while the second should pay the amount that would be paid by two unrelated people each earning $40,000.

In a progressive tax system, that implies the family with children should pay less tax in total than that without.

Clearly there are many different ways in which a tax/transfer system can be structured to achieve horizontal equity. Some countries use income tax scales that are based on "equivilised" income, which involves translating family income into an income per person that corrects for the savings families can secure by sharing costs such as housing; other countries, including Australia, take little account of family size in the income tax structure but make compensatory payments through the welfare system. However, precisely because our system stringently means tests those payments, they do little to ensure horizontal equity at middle income levels. Middle-income families with children therefore face far higher tax rates on per capita income than do families without.

That having children is a choice in no way justifies that result. Even putting aside the principle that all people - young or old - are entitled to horizontal equity, the fact that outcomes reflect choices does not in itself determine their tax consequences. It is, for example, a choice to work hard and earn $120,000, instead of enjoying life and earning half that amount. Yet our tax system extracts a much higher tax take from the hard worker than from those who choose to earn less.

It seems perverse to accept that differentiation with respect to work effort but not to children, all the more so given society's interest in future generations.

There is consequently an argument that we make too little, rather than too much, provision for the difference in capacity to pay between families with children and those without. That hardly means we should simply increase benefits; rather, it means we need a debate based on facts and logic, not slogans and cardboard effigies. Despite all the noise about "middle-class welfare", that debate remains to be had.